‘All the game is in the few’: Why (most) stocks don’t outperform T-Bills
“When I was compiling this I almost didn’t write it up,” Hendrik “Hank” Bessembinder (pictured), a finance professor at Arizona State University, told the Active Advantage conference on Monday (February 27). “I thought that people must know this. Because it’s not exactly rocket science; you take the same database that people have been looking at and actually compound the returns, or convert things to dollar terms. It seems that a lot of people didn’t know. One of the most puzzling questions is why people didn’t know.”
Bessembinder was at the Sydney conference – held by MFS, Orbis and Baillie Gifford – talking new developments to a five-year old piece of research with the provocative title Do Stocks Outperform Treasury Bills? (the first time a finance professor has had much success with marketing, Bessembinder joked, admitting that his paper likely would not have been downloaded some 41,000 times if he’d given it a title like Stock Returns are Skewed).
The global iteration of the research from 2019 shows that the vast majority of stocks do not outperform a Treasury Bill. From 1990 to 2020, the top 1526 firms in the world (out of 63,785 in the global study, so 2.39 per cent), account for 100 per cent of all wealth generation. The remaining 62,259 collectively matched the return on a Treasury Bill; 39.9 per cent generated modest positive wealth, just offsetting the wealth destruction wrought by the remainder.
“The most likely outcome for a stock in the long-run is a loss of essentially 100 per cent,” Bessembinder said. “Over (the 30-year period) the public stock markets were incredible generators of wealth; over $75 trillion in wealth creation. But it was very concentrated – the top five firms generated more than 10 per cent.”
The new dimension to Bessembinder’s research is partially explaining why those top stocks outperform. One possibility is that it’s explained completely by fundamentals. Bessembinder took US stocks from 1970 to 2020 and assessed their decade-horizon returns with five explanatory variables: whether it was a ‘good’ decade for markets; how much the company grew its income; how much it grew its assets; how much it grew its sales; and its average income/assets ratio. Those five variables explain 29.43 per cent of variation in decade horizon stock returns, with income growth the largest contributor.
“It’s interesting that having considered everything that could possibly be considered here, and some of the more tangible things, of some of the things we can readily measure it’s literally the bottom line that’s most important, as well as growth in income,” Bessembinder said.
Another possible explanation is superstar CEOs. When Bessembinder put the question to the data, with top performing firms with more than one CEO assessed against variables of how the markets turned out during their tenure, as well as firm and CEO identity, they collectively explained 19 per cent of the variation in compound annual returns (Bessembinder notes that 80 per cent remains unexplained).
“Within the 20 per cent that is explained, it’s important to know how the year turned out; it’s important to know which firm it is,” Bessembinder said. “Being Apple is different to being Exxon-Mobil, even though they’re both high on the list. The most important thing to hear is that even when we control for whether you were an Apple or a Microsoft the CEO identity is more important than firm identity.”
“Some of you might say ‘isn’t that a bit obvious?’…But I do think management matters a lot in a broad sense and in ways that might be hard to quantify but in any case the data backs it up: the individual CEO matters a lot for long-term performance.”
There are several implications from all this. The first is that public equities are a lot more like venture capital than most people realise.
“If I told you about an asset class where most investments lose money, the most common outcome is that you lose all your money, but there’s a few really big winners that make the whole asset class worthwhile, the response that ‘We already knew that about venture capital’ would be a perfectly reasonable response,” Bessembinder said. “Except that I’m describing public equity investing.”
The textbooks lay out all the reasons why people should have broadly diversified low cost portfolios, Bessembinder said; his study backs that up “and some more”’. If you’re just picking stocks at random, the odds are worse than 50/50 that you’ll underperform the benchmark. But if you can pick those winners and “all the game is in the few, shouldn’t you be striving for the few?”.
“The irony is that when people read my study it was like a Rorschach test,” Bessembinder said. “Some people look at my study and say they should be indexed; others look at it and say they should be actively managed. But there’s ammunition here for both sides that wasn’t there before.”